Automated expense invoice data extraction and processing using OASIS AP
In 2018, a large Australian retailer approached eCloud to look at ways to automate their expense invoices. They had an expense processing system in place but were looking at replacing it due to the inability of their existing provider to deliver the promised solution. The multiple problems encountered with their service led to an overall productivity decline and low staff morale.
The client was looking at resolving a range of issues and to reduce the process time which would take their team over 20 hours to process 18-20,000 expense invoices.
The client had invested in a solution which had not delivered all that it was promised, and was in the process of cancelling the project when their provider recognised that they could not deliver as specified.
- The system would freeze up to 14 times a day and forced the user to log back in each time that happened
- The scanning process would not be able to find the store number is it was in a place other than the specified area on the invoice
- If there was no store number on the invoice, their system could not look at the delivery address to automatically identify the store therefore requiring manual intervention
- It was impossible to use a default store code if no code was found
- It was impossible to isolate the number of scanned invoices to track daily statistics. This led to not being able to measure the success rate of invoices processed without human intervention
- The staff could not search for a specific invoice
- The approval process was very cumbersome for Store Managers
OASIS is an intelligent automated business processing service that provides a fully outsourced – or significantly enhanced in-house – solution for automated scanning, data extraction, purchase order matching, approval workflows which can export approved invoices directly into a customer’s accounts payable system ready for payment.
OASIS uses a combination of optical character recognition, artificial intelligence (image recognition, natural language processing, machine learning), and electronic quality assurance checking processes to ensure the highest level of accuracy in the analysis and processing of invoices regardless of their format. Cloud-based, OASIS was already set to operate as a filter between the sending of invoices and the extraction to the finance system and to ensure the intelligent processing extracts the approved invoices directly to their Oracle finance system with little to no manual intervention.
The solution required special consideration for change management. The Store Managers were already under pressure with a number of system changes implemented throughout the store regarding finance management. eCloud was specifically asked to look at implementing and allow for training and change management with AP users but ensure that no impact or training was required with the Store Managers.
This was done by providing the service and automated analysis and matching prior to the expense invoice being sent to the finance system. This allowed for the store users to simply continue accessing the approval process within their usual finance environment with no knowledge of OASIS AP automated processing activities.
With the solution already in place about to be canceled, the timeline for delivery was extremely sensitive and the solution needed to be in place within a two month period, and it was. It was then followed by a couple of months of active monitoring by the client and eCloud to identify and make adjustments if and when required. Finally, with an IT team already scheduled for other projects, the new project was to run all integration tasks as efficiently as possible.
eCloud’s existing knowledge of the finance solution in place at this large Australian retailer only required a couple of hours from the client’s IT team to get the full workflow running.
Following the implementation of OASIS AP for the automated processing of expense invoices, the client achieved the following results:
- Responsive and reliable system
- True automation: 85% + expense invoices data checked and resolved without exceptions, 15% possible exceptions for one-off or new suppliers or invoice calculation errors
- Automatic electronic transfer of the data into the Finance system for Store Managers to finalise the process
- Ability to detect and cater for GST and non-GST transactions
- Ability to open tasks by Queue or open tasks by the vendor
- The ability of the system to learn as it goes and apply new business rules
- The ability of the AP support team to set new business rules
- Ability to report on and provide statistics on volumes, resolution rates, and other productivity indicators as set by the AP team
- Ability to search for invoices
The client’s team is spending considerably fewer hours to process expense invoices. Following the successful implementation of the expense invoice automation solution, eCloud was asked to implement the same solution at the NZ branch and to provide accounts payable automation services to the retailer’s global head office.